Pharmacological studies on the biological action of cocaine have identified several sites of specific binding in the central nervous system associated with toxicity, self administration behavior, and other features of drug abuse. There are in vitro assays showing good correlation with physiologica effects, whereby experimentalists have measured specific binding constants to well-defined CNS receptors for a series of cocaine derivatives. Since these assays are typically done with brain homogenates, and the receptors are probably membrane bound proteins, there is little hope for a detailed X-ray crystal structure of the receptor, and we must make do with just the binding data. Can we use this limited information to deduce a detailed pic- ture of the shape of the receptor site and the energetics of drug-receptor interactions? How much detail can be deduced? What can we learn if the assa actually involves more than one type of receptor? Can this information be used to suggest cocaine antagonists or drugs with greater specificity for particular receptors? We propose to enhance our rigorous computer algorithm for receptor mapping and apply them to the available experimental data on cocaine binding to CNS receptors, in order to answer these questions.